As machine learning and artificial intelligence advances, the fear of technology taking over human jobs increases. A common concern is that much of jobs that required human labor will be replaced by machines, and many people will lose their usefulness in the market. However, in Ghost Works, the sentiment is that as algorithms make certain tasks obsolete, they also create a need for many new tasks that require humans. These tasks are often done internationally, crowdsourced, and fast. Users of technology assume their applications are powered digitally alone, but in fact, many softwares employ this man powered task force in the background to fulfill what artificial intelligence is unable to accomplish. Because of the lack of awareness, and often intentional hiding, of these workers and these gigs, they are referred to as “ghost work.” These labors are being undervalued because of this opacity in their worth and even existence, as a result, ghost workers are underpaid, leading to a growing pool of invisible underclass.
Though I agree with the idea that new technology will replace antiqued jobs but create new opportunities, what caught my attention while reading was the notion that humans are fundamentally more complex, and different, than artificial intelligence. I disagree with this in a biological and philosophical sense. Most research in psychology and HCI nowadays suggest that humans are better at pattern recognition. Machines, on the other hand, have far superior long term memory storage. Given proper care, memories stored on a drive will not change as time passed, but human memories will fade and be altered by perception and new information. I feel that it is theoretically possible to create an artificial brain that is as powerful as the human brain. However, I agree that human labor may be still necessary on another basis. I think it would be more cost, time, and labor efficient. In a somewhat dystopian view, to create a capable pattern recognition machine through humans, it takes two human laborers and about 17 years of upbringing, costing $233,610, according to data from USDA. This new human will not only be capable of recognizing the difference between hairless puppies and “dick pics”, as the example from the book, but also a wide variety of other tasks.
Back to the main topic of the reading, I think the main concern is that, as of now, there is a large amount of workers being underpaid for labor that will only grow to be more significant and prevalent. If this trend continues, and the ghost workers continue to be invisible to the general public, more people will lose their jobs as artificial intelligence replaces them without realizing new needs are created elsewhere. Simultaneously, the lack of recognition for the value of these new, machine irreplaceable tasks will cause them to be underpaid and under desired. As of now, demand for these works is much less than the abundant supply of workers all over the globe. However, demand is rapidly growing as machine learning and artificial intelligence advances.
How do the ideas of this book conflict with the theory that advancing technology may reduce the need for “work” and undesirable tasks and increase opportunities for humans to pursue creative activities?
Much of ghost work is considered “mindless” gigs. How can we make these tasks less mundane and undesirable?
Can we apply the techniques used to make ghost work “easy” or mindless, for example segmenting tasks and crowdsourcing, to other works to make them easier as well?
Have ghost work always existed or only recently came into being with the rise of AI? What tasks in the past were underpaid and opaque? How are they similar or different?
